HHS Public Access Author manuscript Author Manuscript
J Behav Med. Author manuscript; available in PMC 2017 April 01. Published in final edited form as: J Behav Med. 2016 April ; 39(2): 355–363. doi:10.1007/s10865-016-9712-3.
The interrelations between spiritual well-being, pain interference and depressive symptoms in patients with multiple sclerosis Sheri A. Nsamenang1, Jameson K. Hirsch2, Raluca Topciu3, Andrew D. Goodman4, and Paul R. Duberstein5 Jameson K. Hirsch: [email protected] 1McMaster
Childrens Hospital, Hamilton Health Sciences, Hamilton, Canada
Author Manuscript
2Department
of Psychology, East Tennessee State University, 420 Rogers Stout Hall, Johnson City, TN 37614, USA
3Centre
for Psychiatry, Wolfson Institute of Preventive Medicine, Barts and the London School of Medicine and Dentistry, London, UK
4Department
of Neurology, University of Rochester Medical Center, Rochester, NY, USA
5Departments
of Psychiatry and Family Medicine, University of Rochester Medical Center, Rochester, NY, USA
Abstract Author Manuscript Author Manuscript
Depressive symptoms are common in individuals with multiple sclerosis (MS), and are frequently exacerbated by pain; however, spiritual well-being may allow persons with MS to more effectively cope with pain-related deficits in physical and role functioning. We explored the associations between spiritual well-being, pain interference and depressive symptoms, assessing each as a potential mediator, in eighty-one patients being treated for MS, who completed selfreport measures: Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being Scale, Pain Effects Scale, and Center for Epidemiologic Studies Depression Scale Revised. At the bivariate level, spiritual well-being and its subscale of meaning and peace were negatively associated with depression and pain interference. In mediation models, depression was not related to pain interference via spiritual well-being, or to spiritual well-being via pain interference. Pain interference was related to depression via spiritual well-being and meaning/peace, and to spiritual well-being and meaning/peace via depressive symptoms. Finally, spiritual well-being and meaning/peace were related to depression via pain interference, and to pain interference via depressive symptoms. For patients with MS, a multi-faceted approach to treatment that includes pain reduction and promotion of spiritual well-being may be beneficial, although amelioration of depression remains a critical task.
Correspondence to: Jameson K. Hirsch, [email protected]. Conflict of interest Sheri A. Nsamenang, Jameson K. Hirsch, Raluca Topciu, Andrew D. Goodman and Paul R. Duberstein declare that they have no conflict of interest. Compliance with ethical standards: Human and animal rights and Informed consent All procedures followed were in accordance with ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000. Informed consent was obtained from all patients for being included in the study.
Nsamenang et al.
Page 2
Author Manuscript
Keywords Spiritual well-being; Pain interference; Depressive symptoms; Multiple sclerosis
Introduction
Author Manuscript
Multiple sclerosis (MS) is a chronic, neurological disorder for which there is no cure, and is estimated to affect between 250,000 and 400,000 individuals in the United States, and 2.1 million people worldwide (Devins & Shnek, 2000; Reingold, 2002). Typically, the disease is characterized by unpredictable periods of symptoms such as visual loss or numbness of the limbs, with intermittent periods of partial or full remission, occurring over one to two decades with constant progression and accumulated disability (McNulty et al., 2004). Many individuals with MS are faced with the prospect of enduring years of persisting and worsening symptoms, including pain and risk for depression and impairment in social, physical and psychological functioning (Osborne et al., 2006). Comorbid pain (affecting 44– 80 % of MS patients) and depression (>50 % of MS patients) are common in patients with multiple sclerosis (Hirsh et al., 2009), with greater levels of pain severity and interference related to greater depressive symptoms (Arnett et al., 2008; Alschuler et al., 2013). Pain frequently disrupts the ability to engage in required daily routines and self-care (Osborne et al., 2006), contributes to poor quality of life (Ehde et al., 2003; Harris et al., 2003) and activity limitations (Hunfeld et al., 2001; Smith et al., 2001). Despite its neurological basis, pain is also associated with psychosocial factors (Gatchel et al., 2007).
Author Manuscript Author Manuscript
One such factor, spiritual well-being, is defined as the benefits occurring from the “way individuals seek and express meaning and purpose and the way they experience their connectedness to the moment, to self, to others, to nature, and to the significant or sacred” (Puchalski et al., 2009), and is conceptualized as having two components: meaningfulness, or a sense of purpose in life, and faith, which refers to a sense of comfort in feeling connected with a higher power (Yanez et al., 2009). A growing body of research suggests that spiritual well-being is related to overall health (Bredle et al., 2011), decreased pain severity and better coping with chronic illness (Irvine et al., 2009; Keefe et al., 2001). The existential components of spiritual well-being (e.g., meaning and peace) are often more robustly associated with good health than traditional religious behaviors (e.g., attending church services) (Bekelman et al., 2007; Hirsch et al., 2014). Despite this, some conflicting results exist; for example, several previous studies have found no association between spiritual well-being and pain severity or interference, suggesting that other factors (e.g., variability in spiritual well-being and pain over time) may contribute to differences noted in their association (Mystakidou et al., 2007; Rippentrop et al., 2005). It is clear that the interrelations between spirituality, pain and depression are complex, with often-contradictory findings and bi-directionality; however, the linkages between these variables have not been previously examined in patients with MS. In general, spiritual wellbeing is related to less perceived illness uncertainty and psychological distress (McNulty et al., 2004), as well as better coping and psychosocial adjustment (Irvine et al., 2009; McNulty et al., 2004), in patients with MS. However, previous research also indicates that
J Behav Med. Author manuscript; available in PMC 2017 April 01.
Nsamenang et al.
Page 3
Author Manuscript
pain is deleteriously associated with engagement in spiritual pursuits (i.e., negative effect of pain on ability to complete social routines; distraction) and is linked to depression risk (e.g., resulting from discomfort, isolation) (de Villiers et al., 2015; Dobratz, 2005). As well, depression is often linked to greater levels of pain (e.g., via catastrophizing, comorbid somatic symptoms) and less motivation or ability to engage in spiritual pursuits (Williams et al., 2003; Haythornthwaite et al., 1991).
Author Manuscript
Thus, the purpose of this study is to examine the interrelations between spiritual well-being, pain interference and depressive symptoms in persons with MS. We hypothesized that, at the bivariate level, spiritual well-being would be negatively, and pain interference positively, associated with depressive symptoms. In exploratory mediation models, we analyzed all potential associations between spiritual well-being, and its subscales, and depressive symptoms and pain interference, using each as a predictor, mediator and outcome variable, in turn.
Methods
Author Manuscript
Participants in this IRB-approved study were volunteers recruited from an outpatient, university hospital-based, MS clinic. Informed consent was obtained by trained research staff, and participants completed questionnaires in written format, in a private assessment room. Our sample consisted of 81 participants, ages 30–75 years old (M = 51.12; SD = 9.6), with more women (n = 64; 83 %) than men. The self-reported race of our sample was primarily White (91.4 %; N = 74), with 4.9 % Other (N = 4), 2.5 % Multiracial (N = 2), and 1.3 % Black (N = 1). Participants reported a mean education level of 15.44 years (SD = 2.6). Approximately 59 % (58.8 %; N = 47.63) reported current engagement in religious practice. Most participants were married (60.5 %), 14.8 % divorced, 11.1 % single and never married, 6.2 % widowed, 2.5 % legally separated, and 4.9 % with no reported marital status. Measures
Author Manuscript
The Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being Scale (FACITSp) (Peterman et al., 2002), was developed to assess the spiritual well-being of patients with chronic illness, and includes two subscales: The Meaning and Peace subscale (Mean = 21.24; SD = 7.06), which measures feelings of inner peace and meaning (8 items; e.g., “I feel a sense of purpose in my life”), and the Faith subscale (Mean = 8.74, SD = 4.60), which measures sense of strength and comfort from one's beliefs (4 items; e.g., “I find strength in my faith or spiritual beliefs”). Items are rated on a 5-point Likert scale ranging from 0 (not at all) to 5 (very much), with higher scores indicating greater spiritual well-being (Total Range = 0–40; faith subscale = 0–20; meaning and peace subscale = 0–40). In a series of two studies, among patients with cancer, the FACIT-SP demonstrated good internal consistency for the total score (α = .87; .86), meaning and peace subscale (α = .81; .81), and faith subscale (α = .88; .86) (Peterman et al., 2002). In our sample, alpha coefficients were as follows: total score (α = .88), meaning and peace subscale (α = .91), and faith subscale (α = .84). Pain interference, or the effect of pain on behavior and mood, was measured by the Pain Effects Scale (PES), which consists of 6 items, and is part of the Multiple Sclerosis Quality
J Behav Med. Author manuscript; available in PMC 2017 April 01.
Nsamenang et al.
Page 4
Author Manuscript
of Life Inventory (MSQLI) (Ritvo et al., 1997). Participants indicated the degree to which pain interfered with their mood, ability to walk or move around, sleep, normal work, recreational activities, and enjoyment in life, on a 5-point Likert scale ranging from 1 (not at all) to 5 = (to an extreme degree); higher scores indicate greater impact of pain interference (range = 6–30). Among MS patients, the PES has demonstrated excellent internal consistency (α = .92) (Osborne et al., 2006), which was reflected in the current sample (α = . 94).
Author Manuscript
The Center for Epidemiologic Studies Depression Scale (CES-D-R 20) was used to assess presence and severity of depressive symptoms (Radloff, 1977). Items (N = 20) are rated on a 4-point Likert scale ranging from 0 (rarely or none of the time) to 3 (most or all of the time), with higher scores indicating greater depressive symptoms (range = 0–60). The CES-D-R 20 has excellent internal reliability (α = .90) among patients with MS (Verdier-Taillefer et al., 2001), as it did in the current study (α = .91). Statistical analyses Bivariate analyses—Pearson's product-moment correlations were used to examine zeroorder associations between spiritual well-being, the faith subscale, the meaning and peace subscale, pain interference, and depressive symptoms; no correlation exceeded the cut-off recommended for multicollinearity (r > .80) (Katz, 2006), except for the relation between spiritual well-being and the meaning and peace subscale which, in mediation models, were assessed independently.
Author Manuscript Author Manuscript
Mediation analyses—Multivariate mediation analyses were conducted, consistent with Preacher and Hayes (2008a), which allow for one independent variable (IV), one dependent variable (DV), and more than one mediator variable. We examined the relation between spiritual well-being, and its subscales, and depression via pain interference; for subscale analyses, the spiritual well-being subscale not entered as the IV was covaried (see Fig. 1). We also examined the following models: relation between spiritual well-being and pain interference via depression; relation between depression and pain interference via spiritual well-being; between depression and spiritual well-being via pain interference; between pain interference and depression via spiritual well-being; and, finally, between pain interference and spiritual well-being via depression. There are several potential pathways of significance in a mediation model. The total effect, or (c) pathway, signifies the direct relationship between the IV and DV without controlling for possible MVs. A direct effect, or the (c′) pathway, refers to the relation between the IV and DV after controlling for the possible effects of a MV. A total indirect effect (ab) refers to the role of all potential MVs on the association between an IV and a DV. Finally a specific indirect effect refers to the role of a particular MV(s) in the relationship between an IV and a DV. When there is a reduction in significance from the total to direct effects (c to c′), mediation is said to occur, with these common terms often applied: partial mediation (coefficient remains significant) and full mediation (coefficient is reduced to non-significance). Compared to Baron and Kenny's (1986) approach to mediation analysis, Preacher and Hayes' (2008a) mediation techniques use bootstrap resampling to calculate more accurate
J Behav Med. Author manuscript; available in PMC 2017 April 01.
Nsamenang et al.
Page 5
Author Manuscript
analysis of indirect effects (Hayes, 2009; Preacher & Hayes, 2008b). Bootstrapping resampling provides an estimate of indirect effects and empirical approximations of the sampling distribution of an indirect effect, by resampling a study sample k times (at least 5000 times; in the current study, 10,000 iterations) and generating confidence intervals (CI's), that permit inferences about the size of the indirect effect (Hayes, 2009); when CI's do not cross zero, the model is deemed significant. The techniques can be used on nonnormally distributed data, and allow for indirect effects without the presence of direct effects. Also, the bootstrapping method permits the detection of effects even when the sample size is small (Preacher & Hayes, 2008a). In all models, age and sex were included as covariates, and the alpha (Type I error) level was set at p
J Behav Med. Author manuscript; available in PMC 2017 April 01. Published in final edited form as: J Behav Med. 2016 April ; 39(2): 355–363. doi:10.1007/s10865-016-9712-3.
The interrelations between spiritual well-being, pain interference and depressive symptoms in patients with multiple sclerosis Sheri A. Nsamenang1, Jameson K. Hirsch2, Raluca Topciu3, Andrew D. Goodman4, and Paul R. Duberstein5 Jameson K. Hirsch: [email protected] 1McMaster
Childrens Hospital, Hamilton Health Sciences, Hamilton, Canada
Author Manuscript
2Department
of Psychology, East Tennessee State University, 420 Rogers Stout Hall, Johnson City, TN 37614, USA
3Centre
for Psychiatry, Wolfson Institute of Preventive Medicine, Barts and the London School of Medicine and Dentistry, London, UK
4Department
of Neurology, University of Rochester Medical Center, Rochester, NY, USA
5Departments
of Psychiatry and Family Medicine, University of Rochester Medical Center, Rochester, NY, USA
Abstract Author Manuscript Author Manuscript
Depressive symptoms are common in individuals with multiple sclerosis (MS), and are frequently exacerbated by pain; however, spiritual well-being may allow persons with MS to more effectively cope with pain-related deficits in physical and role functioning. We explored the associations between spiritual well-being, pain interference and depressive symptoms, assessing each as a potential mediator, in eighty-one patients being treated for MS, who completed selfreport measures: Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being Scale, Pain Effects Scale, and Center for Epidemiologic Studies Depression Scale Revised. At the bivariate level, spiritual well-being and its subscale of meaning and peace were negatively associated with depression and pain interference. In mediation models, depression was not related to pain interference via spiritual well-being, or to spiritual well-being via pain interference. Pain interference was related to depression via spiritual well-being and meaning/peace, and to spiritual well-being and meaning/peace via depressive symptoms. Finally, spiritual well-being and meaning/peace were related to depression via pain interference, and to pain interference via depressive symptoms. For patients with MS, a multi-faceted approach to treatment that includes pain reduction and promotion of spiritual well-being may be beneficial, although amelioration of depression remains a critical task.
Correspondence to: Jameson K. Hirsch, [email protected]. Conflict of interest Sheri A. Nsamenang, Jameson K. Hirsch, Raluca Topciu, Andrew D. Goodman and Paul R. Duberstein declare that they have no conflict of interest. Compliance with ethical standards: Human and animal rights and Informed consent All procedures followed were in accordance with ethical standards of the responsible committee on human experimentation (institutional and national) and with the Helsinki Declaration of 1975, as revised in 2000. Informed consent was obtained from all patients for being included in the study.
Nsamenang et al.
Page 2
Author Manuscript
Keywords Spiritual well-being; Pain interference; Depressive symptoms; Multiple sclerosis
Introduction
Author Manuscript
Multiple sclerosis (MS) is a chronic, neurological disorder for which there is no cure, and is estimated to affect between 250,000 and 400,000 individuals in the United States, and 2.1 million people worldwide (Devins & Shnek, 2000; Reingold, 2002). Typically, the disease is characterized by unpredictable periods of symptoms such as visual loss or numbness of the limbs, with intermittent periods of partial or full remission, occurring over one to two decades with constant progression and accumulated disability (McNulty et al., 2004). Many individuals with MS are faced with the prospect of enduring years of persisting and worsening symptoms, including pain and risk for depression and impairment in social, physical and psychological functioning (Osborne et al., 2006). Comorbid pain (affecting 44– 80 % of MS patients) and depression (>50 % of MS patients) are common in patients with multiple sclerosis (Hirsh et al., 2009), with greater levels of pain severity and interference related to greater depressive symptoms (Arnett et al., 2008; Alschuler et al., 2013). Pain frequently disrupts the ability to engage in required daily routines and self-care (Osborne et al., 2006), contributes to poor quality of life (Ehde et al., 2003; Harris et al., 2003) and activity limitations (Hunfeld et al., 2001; Smith et al., 2001). Despite its neurological basis, pain is also associated with psychosocial factors (Gatchel et al., 2007).
Author Manuscript Author Manuscript
One such factor, spiritual well-being, is defined as the benefits occurring from the “way individuals seek and express meaning and purpose and the way they experience their connectedness to the moment, to self, to others, to nature, and to the significant or sacred” (Puchalski et al., 2009), and is conceptualized as having two components: meaningfulness, or a sense of purpose in life, and faith, which refers to a sense of comfort in feeling connected with a higher power (Yanez et al., 2009). A growing body of research suggests that spiritual well-being is related to overall health (Bredle et al., 2011), decreased pain severity and better coping with chronic illness (Irvine et al., 2009; Keefe et al., 2001). The existential components of spiritual well-being (e.g., meaning and peace) are often more robustly associated with good health than traditional religious behaviors (e.g., attending church services) (Bekelman et al., 2007; Hirsch et al., 2014). Despite this, some conflicting results exist; for example, several previous studies have found no association between spiritual well-being and pain severity or interference, suggesting that other factors (e.g., variability in spiritual well-being and pain over time) may contribute to differences noted in their association (Mystakidou et al., 2007; Rippentrop et al., 2005). It is clear that the interrelations between spirituality, pain and depression are complex, with often-contradictory findings and bi-directionality; however, the linkages between these variables have not been previously examined in patients with MS. In general, spiritual wellbeing is related to less perceived illness uncertainty and psychological distress (McNulty et al., 2004), as well as better coping and psychosocial adjustment (Irvine et al., 2009; McNulty et al., 2004), in patients with MS. However, previous research also indicates that
J Behav Med. Author manuscript; available in PMC 2017 April 01.
Nsamenang et al.
Page 3
Author Manuscript
pain is deleteriously associated with engagement in spiritual pursuits (i.e., negative effect of pain on ability to complete social routines; distraction) and is linked to depression risk (e.g., resulting from discomfort, isolation) (de Villiers et al., 2015; Dobratz, 2005). As well, depression is often linked to greater levels of pain (e.g., via catastrophizing, comorbid somatic symptoms) and less motivation or ability to engage in spiritual pursuits (Williams et al., 2003; Haythornthwaite et al., 1991).
Author Manuscript
Thus, the purpose of this study is to examine the interrelations between spiritual well-being, pain interference and depressive symptoms in persons with MS. We hypothesized that, at the bivariate level, spiritual well-being would be negatively, and pain interference positively, associated with depressive symptoms. In exploratory mediation models, we analyzed all potential associations between spiritual well-being, and its subscales, and depressive symptoms and pain interference, using each as a predictor, mediator and outcome variable, in turn.
Methods
Author Manuscript
Participants in this IRB-approved study were volunteers recruited from an outpatient, university hospital-based, MS clinic. Informed consent was obtained by trained research staff, and participants completed questionnaires in written format, in a private assessment room. Our sample consisted of 81 participants, ages 30–75 years old (M = 51.12; SD = 9.6), with more women (n = 64; 83 %) than men. The self-reported race of our sample was primarily White (91.4 %; N = 74), with 4.9 % Other (N = 4), 2.5 % Multiracial (N = 2), and 1.3 % Black (N = 1). Participants reported a mean education level of 15.44 years (SD = 2.6). Approximately 59 % (58.8 %; N = 47.63) reported current engagement in religious practice. Most participants were married (60.5 %), 14.8 % divorced, 11.1 % single and never married, 6.2 % widowed, 2.5 % legally separated, and 4.9 % with no reported marital status. Measures
Author Manuscript
The Functional Assessment of Chronic Illness Therapy-Spiritual Well-Being Scale (FACITSp) (Peterman et al., 2002), was developed to assess the spiritual well-being of patients with chronic illness, and includes two subscales: The Meaning and Peace subscale (Mean = 21.24; SD = 7.06), which measures feelings of inner peace and meaning (8 items; e.g., “I feel a sense of purpose in my life”), and the Faith subscale (Mean = 8.74, SD = 4.60), which measures sense of strength and comfort from one's beliefs (4 items; e.g., “I find strength in my faith or spiritual beliefs”). Items are rated on a 5-point Likert scale ranging from 0 (not at all) to 5 (very much), with higher scores indicating greater spiritual well-being (Total Range = 0–40; faith subscale = 0–20; meaning and peace subscale = 0–40). In a series of two studies, among patients with cancer, the FACIT-SP demonstrated good internal consistency for the total score (α = .87; .86), meaning and peace subscale (α = .81; .81), and faith subscale (α = .88; .86) (Peterman et al., 2002). In our sample, alpha coefficients were as follows: total score (α = .88), meaning and peace subscale (α = .91), and faith subscale (α = .84). Pain interference, or the effect of pain on behavior and mood, was measured by the Pain Effects Scale (PES), which consists of 6 items, and is part of the Multiple Sclerosis Quality
J Behav Med. Author manuscript; available in PMC 2017 April 01.
Nsamenang et al.
Page 4
Author Manuscript
of Life Inventory (MSQLI) (Ritvo et al., 1997). Participants indicated the degree to which pain interfered with their mood, ability to walk or move around, sleep, normal work, recreational activities, and enjoyment in life, on a 5-point Likert scale ranging from 1 (not at all) to 5 = (to an extreme degree); higher scores indicate greater impact of pain interference (range = 6–30). Among MS patients, the PES has demonstrated excellent internal consistency (α = .92) (Osborne et al., 2006), which was reflected in the current sample (α = . 94).
Author Manuscript
The Center for Epidemiologic Studies Depression Scale (CES-D-R 20) was used to assess presence and severity of depressive symptoms (Radloff, 1977). Items (N = 20) are rated on a 4-point Likert scale ranging from 0 (rarely or none of the time) to 3 (most or all of the time), with higher scores indicating greater depressive symptoms (range = 0–60). The CES-D-R 20 has excellent internal reliability (α = .90) among patients with MS (Verdier-Taillefer et al., 2001), as it did in the current study (α = .91). Statistical analyses Bivariate analyses—Pearson's product-moment correlations were used to examine zeroorder associations between spiritual well-being, the faith subscale, the meaning and peace subscale, pain interference, and depressive symptoms; no correlation exceeded the cut-off recommended for multicollinearity (r > .80) (Katz, 2006), except for the relation between spiritual well-being and the meaning and peace subscale which, in mediation models, were assessed independently.
Author Manuscript Author Manuscript
Mediation analyses—Multivariate mediation analyses were conducted, consistent with Preacher and Hayes (2008a), which allow for one independent variable (IV), one dependent variable (DV), and more than one mediator variable. We examined the relation between spiritual well-being, and its subscales, and depression via pain interference; for subscale analyses, the spiritual well-being subscale not entered as the IV was covaried (see Fig. 1). We also examined the following models: relation between spiritual well-being and pain interference via depression; relation between depression and pain interference via spiritual well-being; between depression and spiritual well-being via pain interference; between pain interference and depression via spiritual well-being; and, finally, between pain interference and spiritual well-being via depression. There are several potential pathways of significance in a mediation model. The total effect, or (c) pathway, signifies the direct relationship between the IV and DV without controlling for possible MVs. A direct effect, or the (c′) pathway, refers to the relation between the IV and DV after controlling for the possible effects of a MV. A total indirect effect (ab) refers to the role of all potential MVs on the association between an IV and a DV. Finally a specific indirect effect refers to the role of a particular MV(s) in the relationship between an IV and a DV. When there is a reduction in significance from the total to direct effects (c to c′), mediation is said to occur, with these common terms often applied: partial mediation (coefficient remains significant) and full mediation (coefficient is reduced to non-significance). Compared to Baron and Kenny's (1986) approach to mediation analysis, Preacher and Hayes' (2008a) mediation techniques use bootstrap resampling to calculate more accurate
J Behav Med. Author manuscript; available in PMC 2017 April 01.
Nsamenang et al.
Page 5
Author Manuscript
analysis of indirect effects (Hayes, 2009; Preacher & Hayes, 2008b). Bootstrapping resampling provides an estimate of indirect effects and empirical approximations of the sampling distribution of an indirect effect, by resampling a study sample k times (at least 5000 times; in the current study, 10,000 iterations) and generating confidence intervals (CI's), that permit inferences about the size of the indirect effect (Hayes, 2009); when CI's do not cross zero, the model is deemed significant. The techniques can be used on nonnormally distributed data, and allow for indirect effects without the presence of direct effects. Also, the bootstrapping method permits the detection of effects even when the sample size is small (Preacher & Hayes, 2008a). In all models, age and sex were included as covariates, and the alpha (Type I error) level was set at p
